基于RGARCH-Copula模型的中美股市尾部相关性研究

Research on Tail Correlation of China and US Stock Markets Based on R-GARCH-Copula Model

本文采用上证综合指数和标普500指数的收益率和已实现测度数据,利用已实现GARCH-Copula(RGARCH-Copula)模型对中美股市尾部相关性进行了研究。具体地,首先利用RGARCH模型对数据进行过滤,然后利用两步估计法(IFM)对各种Copula模型进行估计,通过对数似然值和两种信息准则的比较,选出最佳Copula模型,进而对两市场尾部相关性进行分析。实证结果表明,SurvivalGumbelCopula表现较好,即两市场的相关性存在非对称特征,下尾相关性较强,不存在上尾相关性,说明美国股市大跌中国股市也有较大的概率会大跌,但美国股市大涨中国股市却不一定会大涨。

The paper applies the realized GARCH-Copula (RGARCH-Copula)model to study the tail correlation of China and US stock markets, with thedata on the yeilds and realized measures of Shanghai Stock Exchange Composite Index and the S&P 500 Index. Specifically, the data arefirst filtered by the RGARCH model, and then the two-step estimation method (IFM) is applied into various Copula models. The best Copula model is selected by comparing the log likelihood valuesand the two information criteria. Then the tail correlation between the two markets is analyzed. The empirical results show that Survival Gumbel Copula performs well, that is, the correlation between the two markets is asymmetric, the correlation between the lower tail is strong, and there is no correlation between the upper tail, indicating that when the US stock market falls sharply, there is a large probability that Chinese stock market also falls abruptly. However, when the US stock market rise sharply, the Chinese stock market will not rise likewise.